PromptArmor 测试多款模型网络攻击风险，称 DeepSeek、Kimi、GLM、Qwen 比…
We tested across multiple cyber risks, including phishing, data exfiltration and propensity to run malware. The differences are shocking! https://t.co/3DpnjZQu4
AI 导读
PromptArmor 称其用多款高传播攻击样本测试了钓鱼、数据外泄和运行恶意软件倾向等网络风险，DeepSeek、Kimi、GLM、Qwen 模型受攻击的可能性比 GPT、Claude、Gemini 高 88%。该结论来自其主推文，具体测试方法和样本规模未在推文中说明。
正文
We tested across multiple cyber risks, including phishing, data exfiltration and propensity to run malware. The differences are shocking! https://t.co/3DpnjZQu4f